An artificial intelligence-based navigation path planning method

Abstract

The application relates to the technical field of artificial intelligence robot route planning, in particular to a navigation path planning method based on artificial intelligence, S100: a global 2d environment is constructed before starting navigation; S200: a scanned 2d composition is subjected to gridding treatment; S300: when a sensor with a proportion a of a robot head and a body part senses that an object is close to and stays for more than a time b, the head and the body are automatically turned to a side where the object is located, and an LCD screen on the body is lighted; S400: exhibits exhibits and locations in a museum; S500: a path is planned according to a destination selected by a tourist; S600: an optimal path is planned through algorithm operation of an evaluation function; and S700: the planned path is displayed, and the tourist is led to the destination, the application improves the A* algorithm, improves the heuristic function, and improves the accuracy of the planned path.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence robot route planning technology, specifically to a navigation path planning method based on artificial intelligence. Background Technology

[0002] Since the implementation of the informatization development strategy, digitalization has permeated all aspects of people's lives. In the field of culture and tourism, large tourist venues such as museums, memorial halls, and cultural centers have successively incorporated digital management programs. Specifically for museums, building digital museums is an inevitable path from the era of traditional exhibitions to the era of emerging technologies, while also giving museums new life and allowing the public to understand the exhibits within the museum from multiple dimensions.

[0003] Besides digitizing museum exhibits, many museums also use navigation robots to provide services to visitors. Traditional navigation robots can be divided into two types based on their movement patterns: line-following robots and autonomous navigation robots. Line-following robots move back and forth by scanning a pre-laid, colored line on the ground using sensors. Their advantage is efficient movement within a limited area, but they are limited to simple, repetitive, and mechanical tasks, and cannot handle the diverse needs of large spaces. Autonomous navigation robots, on the other hand, can navigate autonomously in unfamiliar and irregular environments to complete tasks. The key technologies involved are localization, map building, and path planning. For path planning, mature global planning algorithms exist, such as Dijkstra's algorithm and A* algorithm. While Dijkstra's algorithm can find the shortest path for users, its breadth-first nature requires exploring a vast area to generate a result, making it very time-consuming. The A* algorithm, a combination of Dijkstra's and greedy algorithms, is significantly faster than Dijkstra's, but its path generation cannot guarantee quality in large and complex environments. Furthermore, the application of algorithms must consider the actual usage environment and specific problems; therefore, there are currently few mature methods in the field of museum navigation robots to meet the increasing demands of visitors. Summary of the Invention

[0004] The purpose of this invention is to provide an artificial intelligence-based navigation path planning method to solve the problems mentioned in the background art.

[0005] To address the aforementioned technical problems, the present invention provides the following technical solution: a navigation path planning method based on artificial intelligence, the navigation path planning method comprising:

[0006] Step S100: Before starting navigation, use a laser scanner and camera to scan the interior environment of the museum to construct a global 2D environment;

[0007] Step S200: Rasterize the scanned 2D map and embed the exhibits and locations into nodes on the grid. The area where the exhibits and locations are located occupies one or more grids, and each grid corresponds to one node.

[0008] Step S300: When the sensor of the robot's head and body parts accounting for a percentage of a senses that an object is approaching and stays for more than time b, the robot automatically turns its head and body to the side where the object is located and lights up the LCD screen on the robot body. Here, a and b are preset values, a is 80% and b is 1.5s.

[0009] Step S400: The exhibits and locations within the museum are displayed to visitors on the screen in the form of a map or list;

[0010] Step S500: After the tourist selects the destination, the robot's location is detected by the inertial navigation sensor, and the route is planned according to the tourist's selected destination.

[0011] Step S600: Perform algorithmic calculations using the evaluation function to plan the optimal path;

[0012] Step S700: After the algorithm explores the target node, the A* algorithm stops searching and begins to backtrack to the center point of the robot's location, displays the planned path on the LCD screen, and leads the tourists to the destination.

[0013] This invention improves the path planning algorithm A* by modifying its heuristic function. This improves the algorithm's search efficiency, enhances the accuracy of the planned paths, and makes the improved algorithm more suitable for museum scenarios.

[0014] Furthermore, step S500 includes:

[0015] Step S501: If the visitor selects a specific product, set the target point at a node next to the exhibit and close to the center of the area.

[0016] Step S502: If the visitor selects a certain location, then set the target point at the node where the entrance of that location is located.

[0017] This invention allows tourists to choose their own destinations, plans routes based on different destinations, selects the optimal route, and makes different choices for the nodes reached by different destinations, making it more convenient and faster, improving the accuracy of the routes, and meeting the different needs of users.

[0018] Furthermore, the algorithm is an improved A* algorithm:

[0019] The evaluation function of the A* algorithm consists of the actual cost function g(n) and the heuristic function h(n), and its function expression is as follows:

[0020] f(n) = g(n) + h(n)

[0021] Where g(n) is the actual cost from the starting node to the current node; h(n) is the estimated cost from the current node to the target node. Distance calculation methods include Manhattan distance and Euclidean distance. Manhattan distance, also known as block distance, is calculated by summing the absolute values ​​of the distances between two points on the coordinate axes; while Euclidean distance, also known as Euclidean distance, is the straight-line distance between two points.

[0022] The first step in the A* algorithm's computation is to rasterize the map, with each square representing a node on the map. The A* algorithm also uses two lists to store nodes: an open list (containing nodes to be expanded) and a closed list (containing expanded nodes). Starting from the origin as the center point, the algorithm calculates g(n) as 0 for the origin and h(n) as the estimated cost from the origin to the destination, placing the result in the closed list. Then, it calculates g(n) and h(n) for the eight nodes surrounding the origin and places them in the open list. Next, it selects the node with the smallest evaluation function value in the open list as the center point, removes it from the open list, and places it in the closed list. It then expands the surrounding points sequentially. If the point to be expanded is already in the open list, it selects the node with the smallest g(n) value. This process is repeated for each subsequent expansion until the target point is found.

[0023] The accuracy and rationality of the heuristic function h(n) have a significant impact on the algorithm's performance. Furthermore, the A* algorithm can get stuck in local optima in some symmetric scenarios; that is, when the current center point is located symmetrically to the target node, the algorithm will search back and forth between the two points, reaching the maximum search depth and increasing search time. Moreover, in the formula f(n) = g(n) + h(n), the weight of the heuristic function in the overall evaluation function also greatly affects the algorithm's final path planning performance, as shown in the following results:

[0024] When h(n) = 0, the function represents that the A* algorithm degenerates into Dijkstra's algorithm, which means that the shortest path can be found, but the search speed is the slowest.

[0025] When h(n) < actual cost, the function means that the smaller h(n) is, the more nodes A* expands, and the slower it runs. The result is that an optimal path can be found, but the running speed is slower.

[0026] When h(n) = actual cost, the function represents the standard A* algorithm, and the result is: the standard path;

[0027] When h(n) > actual cost, the function indicates that fewer nodes are searched, meaning that the operation is faster, but it is not necessarily the optimal path.

[0028] When h(n) >> g(n), the function representation approximates a greedy algorithm, indicating that while it achieves the fastest computation, the path may not be optimal. The improved heuristic function h(n) is expressed as follows:

[0029] h(n) = ω1D + ω2r;

[0030] Where D is the estimated cost from the current node to the target node, ω1 is the weight controlling D, r is a random number between 0 and 0.5, and ω2 is the weight controlling r.

[0031] Furthermore, the assignments of ω1, ω2, and r in the function expression of the heuristic function h(n) are as follows:

[0032]

[0033] Where N is the maximum distance from the starting point to the target point.

[0034]

[0035] The farther the current node is from the target node, the larger ω1 becomes, and the greater the weight of h(n) in the evaluation function. The search algorithm then tends to search fewer nodes to quickly find a path. Conversely, when the distance is closer, the search algorithm tends to search more surrounding nodes to ensure the best route to the destination is found.

[0036] In summary, the improved evaluation function formula is as follows:

[0037] f(n) = g(n) + ω1D + ω2r;

[0038] Where g(n) is the actual cost function, and ω1D+ω2r is the heuristic function;

[0039] ω1=exp(ε*D / N)-1,ε>0;

[0040]

[0041] D is the distance from the current node to the target node;

[0042] N is the maximum distance from the starting point to the target point;

[0043]

[0044] r is a random number between 0 and 0.5.

[0045] The x and y coordinates are established on a coordinate system based on two perpendicular axes of the 2D drawing.

[0046] This invention improves the path planning algorithm A* by modifying its heuristic function, thereby mitigating the impact of the inaccuracy of the heuristic function and the problem of easily getting trapped in local optima.

[0047] Furthermore, in step S100, the scanning unit scans the internal environment of the museum. The scanning unit consists of a laser scanner and a camera. The 2D environment construction unit constructs a global 2D environment based on the scanning results of the scanning unit.

[0048] This invention scans the interior environment of a museum and creates a global 2D environment, providing users with a more detailed and accurate representation of the environment, enabling visitors to better experience and understand the exhibits.

[0049] Furthermore, in step S200, the grid unit performs rasterization processing on the 2D composition. The rasterization processing method is to use Adobe Photoshop to convert the vector graphics and non-pixel images of the 2D composition into pixel images. The embedding unit is used to embed exhibits and locations into the nodes of the grid. The embedding method is to use Adobe Photoshop to create images of exhibits and locations, add them to a new layer, and align them with the nodes.

[0050] This invention performs rasterization processing on 2D compositions, converting vector graphics and non-pixel images into pixel images, thereby enhancing image quality, improving image processing efficiency, and supporting visualization analysis.

[0051] Furthermore, in step S300, the sensor unit senses the time it takes for an object to approach and remain near the robot. When the time exceeds 1.5 seconds, the robot automatically turns its head and body to the side where the object is located, and the communication unit is used to turn on the LCD screen.

[0052] This invention detects the time it takes for an object to approach and remain near the robot, automatically turns the robot around, and lights up the LCD screen, making it more convenient for tourists to use the robot.

[0053] Furthermore, in step S400, the visualization unit displays the exhibits and locations within the museum to visitors on the screen in the form of a map or a list.

[0054] This invention presents exhibits and locations within a museum to visitors in the form of a map or list, enabling visitors to have a clearer understanding of the distribution of exhibits and locations within the museum.

[0055] Furthermore, in step S500, the selection unit is used for tourists to select the destination they want to reach. The destination includes exhibits and locations within the museum. The sensor detection unit detects the location of the robot. The decision unit plans a path based on the destination selected by the user in the selection unit. The path planning method is as follows: when the tourist selects a specific product, the target point is set at a node next to the exhibit and close to the center of the area; when the tourist selects a location, the target point is set at the node where the entrance to that location is located.

[0056] This invention improves the tourist experience by planning the optimal route to the destination chosen by tourists, enabling them to reach their destination in the shortest time and making it more convenient.

[0057] Furthermore, in step S600, the A* algorithm unit plans the path according to the evaluation function, selects the fastest path, and ensures that the route is optimal.

[0058] This invention improves the A* algorithm and enhances the heuristic function, thereby increasing the accuracy of the planned path.

[0059] Furthermore, in step S700, the reporting unit reports and displays the planned path, and the robot immediately leads the tourists to their destination.

[0060] This invention displays the planned path and uses a robot to guide tourists to their destination, enabling tourists to clearly understand the path they are taking.

[0061] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0062] (1) This invention improves the path planning algorithm A* by modifying its heuristic function, thereby improving the effects of the inaccuracy of the heuristic function of the A* algorithm and the problem of easily getting trapped in local optima. This improves the algorithm's search efficiency, enhances the accuracy of the planned path, and makes the improved algorithm more suitable for the needs of museum scenarios.

[0063] (2) This invention allows tourists to choose their own destination, plan the route according to different destinations, select the optimal route, and make different selections for the nodes reached by different destinations, which is more convenient and faster, improves the accuracy of the route, and meets the different usage needs of users.

[0064] (3) This invention scans the interior environment of the museum and establishes a global 2D environment, so that users have a more detailed and accurate representation of the environment, and visitors can better experience and understand the exhibits.

[0065] (4) The present invention performs rasterization processing on 2D composition, converts vector and non-pixel images into pixel images, enhances image quality, improves image processing efficiency, and supports visualization analysis. Attached Figure Description

[0066] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0067] Figure 1 This is a flowchart illustrating an artificial intelligence-based navigation path planning method according to the present invention.

[0068] Figure 2 This is a schematic diagram of the structure of a navigation path planning method based on artificial intelligence according to the present invention. Detailed Implementation

[0069] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0070] Please see Figure 1 The present invention provides the following technical solution:

[0071] Step S100: Before starting navigation, use a laser scanner and camera to scan the interior environment of the museum to construct a global 2D environment;

[0072] Step S200: Rasterize the scanned 2D map and embed the exhibits and locations into nodes on the grid. The area where the exhibits and locations are located occupies one or more grids, and each grid corresponds to one node.

[0073] Step S300: When the sensor of the robot's head and body parts accounting for a percentage of a senses that an object is approaching and stays for more than time b, the robot automatically turns its head and body to the side where the object is located and lights up the LCD screen on the robot body. Here, a and b are preset values, a is 80% and b is 1.5s.

[0074] Step S400: The exhibits and locations within the museum are displayed to visitors on the screen in the form of a map or list;

[0075] Step S500: After the tourist selects the destination, the robot's location is detected by the inertial navigation sensor, and the route is planned according to the tourist's selected destination.

[0076] Step S600: Perform algorithmic calculations using the evaluation function to plan the optimal path;

[0077] Step S700: After the algorithm explores the target node, the A* algorithm stops searching and begins to backtrack to the center point of the robot's location, displays the planned path on the LCD screen, and leads the tourists to the destination.

[0078] This invention improves the path planning algorithm A* by modifying its heuristic function. This improves the algorithm's search efficiency, enhances the accuracy of the planned paths, and makes the improved algorithm more suitable for museum scenarios.

[0079] Step S500 includes:

[0080] Step S501: If the visitor selects a specific product, the target point is set at a node next to the exhibit and close to the center of the area.

[0081] Step S502: If the visitor selects a location, then set the target point at the node where the entrance of that location is located.

[0082] This invention allows tourists to choose their own destinations, plans routes based on different destinations, selects the optimal route, and makes different choices for the nodes reached by different destinations, making it more convenient and faster, improving the accuracy of the routes, and meeting the different needs of users.

[0083] The algorithm described is an improved A* algorithm:

[0084] The evaluation function of the A* algorithm consists of the actual cost function g(n) and the heuristic function h(n), and its function expression is as follows:

[0085] f(n) = g(n) + h(n);

[0086] Where g(n) is the actual cost from the starting node to the current node; h(n) is the estimated cost from the current node to the target node. Distance calculation methods include Manhattan distance and Euclidean distance. Manhattan distance, also known as block distance, is calculated by summing the absolute values ​​of the distances between two points on the coordinate axes; while Euclidean distance, also known as Euclidean distance, is the straight-line distance between two points.

[0087] The first step in the A* algorithm's computation is to rasterize the map, with each square representing a node on the map. The A* algorithm also uses two lists to store nodes: an open list (containing nodes to be expanded) and a closed list (containing expanded nodes). Starting from the origin as the center point, the algorithm calculates g(n) as 0 for the origin and h(n) as the estimated cost from the origin to the destination, placing the result in the closed list. Then, it calculates g(n) and h(n) for the eight nodes surrounding the origin and places them in the open list. Next, it selects the node with the smallest evaluation function value in the open list as the center point, removes it from the open list, and places it in the closed list. It then expands the surrounding points sequentially. If the point to be expanded is already in the open list, it selects the node with the smallest g(n) value. This process is repeated for each subsequent expansion until the target point is found.

[0088] The accuracy and rationality of the heuristic function h(n) have a significant impact on the algorithm's performance. Furthermore, the A* algorithm can get stuck in local optima in some symmetric scenarios; that is, when the current center point is located symmetrically to the target node, the algorithm will search back and forth between the two points, reaching the maximum search depth and increasing search time. Moreover, in the formula f(n) = g(n) + h(n), the weight of the heuristic function in the overall evaluation function also greatly affects the algorithm's final path planning performance, as shown in the following results:

[0089] When h(n) = 0, the function represents that the A* algorithm degenerates into Dijkstra's algorithm, which means that the shortest path can be found, but the search speed is the slowest.

[0090] When h(n) < actual cost, the function means that the smaller h(n) is, the more nodes A* expands, and the slower it runs. The result is that an optimal path can be found, but the running speed is slower.

[0091] When h(n) = actual cost, the function represents the standard A* algorithm, and the result is: the standard path;

[0092] When h(n) > actual cost, the function indicates that fewer nodes are searched, meaning that the operation is faster, but it is not necessarily the optimal path.

[0093] When h(n) >> g(n), the function representation approximates a greedy algorithm, indicating that while it achieves the fastest computation, the path may not be optimal. The improved heuristic function h(n) is expressed as follows:

[0094] h(n) = ω1D + ω2r;

[0095] Where D is the estimated cost from the current node to the target node, ω1 is the weight controlling D, r is a random number between 0 and 0.5, and ω2 is the weight controlling r.

[0096] Furthermore, the assignments of ω1, ω2, and r in the function expression of the heuristic function h(n) are as follows:

[0097]

[0098] Where N is the maximum distance from the starting point to the target point.

[0099]

[0100] The farther the current node is from the target node, the larger ω1 becomes, and the greater the weight of h(n) in the evaluation function. The search algorithm then tends to search fewer nodes to quickly find a path. Conversely, when the distance is closer, the search algorithm tends to search more surrounding nodes to ensure the best route to the destination is found.

[0101] In summary, the improved evaluation function formula is as follows:

[0102] f(n) = g(n) + ω1D + ω2r;

[0103] Where g(n) is the actual cost function, and ω1D+ω2r is the heuristic function;

[0104] ω1=exp(ε*D / N)-1,ε>0;

[0105]

[0106] D is the distance from the current node to the target node;

[0107] N is the maximum distance from the starting point to the target point;

[0108]

[0109] r is a random number between 0 and 0.5.

[0110] The x and y coordinates are established on a coordinate system based on two perpendicular axes of the 2D drawing.

[0111] This invention improves the path planning algorithm A* by modifying its heuristic function, thereby mitigating the impact of the inaccuracy of the heuristic function and the problem of easily getting trapped in local optima.

[0112] Assuming this is applied to a small museum, i.e., ε is 4, and further assuming N = 10, and D takes values ​​of 0, 2, 4, 6, 8, and 10, we can obtain the following:

[0113] When D = 0, the rounded ω1 is 0;

[0114] When D = 2, the rounded ω1 is 1;

[0115] When D = 4, the rounded ω1 is 4;

[0116] When D = 6, the rounded ω1 is 10;

[0117] When D = 8, the rounded ω1 is 24;

[0118] When D = 10, the rounded ω1 is 54;

[0119] Please see Figure 2 The structure of an AI-based navigation path planning method:

[0120] In step S100, the scanning unit scans the interior environment of the museum. The scanning unit consists of a laser scanner and a camera. The 2D environment construction unit constructs a global 2D environment based on the scanning results of the scanning unit.

[0121] This invention scans the interior environment of a museum and creates a global 2D environment, providing users with a more detailed and accurate representation of the environment, enabling visitors to better experience and understand the exhibits.

[0122] In step S200, the raster unit rasterizes the 2D composition. The rasterization method is to use Adobe Photoshop to convert the vector graphics and non-pixel images of the 2D composition into pixel images. The embedding unit is used to embed exhibits and locations into the nodes of the raster. The embedding method is to create images of exhibits and locations using Adobe Photoshop, add them to a new layer, and align them with the nodes.

[0123] This invention performs rasterization processing on 2D compositions, converting vector graphics and non-pixel images into pixel images, thereby enhancing image quality, improving image processing efficiency, and supporting visualization analysis.

[0124] In step S300, the sensor unit senses the time it takes for an object to approach and remain near the robot. When the time exceeds 1.5 seconds, the robot automatically turns its head and body to the side where the object is located, and the communication unit is used to turn on the LCD screen.

[0125] This invention detects the time it takes for an object to approach and remain near the robot, automatically turns the robot around, and lights up the LCD screen, making it more convenient for tourists to use the robot.

[0126] In step S400, the visualization unit displays the exhibits and locations within the museum to visitors on the screen in the form of a map or a list.

[0127] This invention presents exhibits and locations within a museum to visitors in the form of a map or list, enabling visitors to have a clearer understanding of the distribution of exhibits and locations within the museum.

[0128] In step S500, the selection unit is used for tourists to select the destination they want to reach. The destination includes exhibits and locations within the museum. The sensor detection unit detects the location of the robot. The decision unit plans a path based on the destination selected by the user in the selection unit. The path planning method is as follows: when the tourist selects a specific product, the target point is set at a node next to the exhibit and close to the center of the area; when the tourist selects a location, the target point is set at the node where the entrance of that location is located.

[0129] This invention improves the tourist experience by planning the optimal route to the destination chosen by tourists, enabling them to reach their destination in the shortest time and making it more convenient.

[0130] In step S600, the A* algorithm unit plans the path according to the evaluation function, selects the fastest path, and ensures that the route is optimal.

[0131] This invention improves the A* algorithm and enhances the heuristic function, thereby increasing the accuracy of the planned path.

[0132] In step S700, the reporting unit reports and displays the planned path, and the robot immediately leads the tourists to their destination.

[0133] This invention displays the planned path and uses a robot to guide tourists to their destination, enabling tourists to clearly understand the path they are taking.

[0134] The working principle of this invention is as follows: Before navigation begins, a laser scanner and camera scan the interior environment of the museum to construct a global 2D environment. The scanned 2D map is rasterized, and exhibits and locations are embedded into nodes on the grid. When 80% of the sensors on the robot's head and body detect an object approaching and remaining there for more than 1.5 seconds, the robot automatically turns its head and body to the side of the object and lights up the LCD screen on its body. The screen displays the exhibits and locations in the museum to the visitor in map or list form. After the visitor selects their desired destination, the robot begins to detect its position using inertial navigation sensors and starts planning a path based on the visitor's chosen destination. An algorithm is performed using an evaluation function. Initially, the fastest path is quickly planned. Gradually, in the later stages, the algorithm tends to explore more surrounding nodes to ensure the optimal route. When the algorithm finds the target node, the A* algorithm stops searching and begins to backtrack to the center point of the robot's current location. The planned path is then displayed on the LCD screen, guiding the visitor to their destination.

[0135] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0136] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A navigation path planning method based on artificial intelligence, characterized in that, The navigation path planning method includes: Step S100: Before starting navigation, use a laser scanner and camera to scan the interior environment of the museum to construct a global 2D environment; Step S200: Rasterize the scanned 2D map and embed the exhibits and locations into nodes on the grid. The area where the exhibits and locations are located occupies one or more grids, and each grid corresponds to one node. Step S300: When the sensor of the robot's head and body parts accounting for a percentage of a senses that an object is approaching and stays for more than time b, the robot automatically turns its head and body to the side where the object is located and lights up the LCD screen on the robot body, where a and b are preset values. Step S400: The exhibits and locations within the museum are displayed to visitors on the screen in the form of a map or list; Step S500: After the tourist selects the destination, the robot's location is detected by the inertial navigation sensor, and the route is planned according to the tourist's selected destination. Step S500 includes: Step S501: If the visitor selects a specific exhibit, then set the target point at the node corresponding to the center point of the exhibit's grid area. Step S502: If the visitor selects a location, set the target point at the node where the entrance of that location is located; Step S600: Perform algorithmic calculations using the evaluation function to plan the optimal path; The algorithm described is an improved version of the A* algorithm: The evaluation function formula for the improved A* algorithm is shown below: f(n)=g(n)+ω1D+ω2r Where g(n) is the actual cost function, and ω1D+ω2r is the heuristic function; ω1=exp(ε*D / N)-1,ε>0; D is the distance from the current node to the target node; N is the maximum distance from the starting point to the target point; r is a random number between 0 and 0.5; Step S700: After the algorithm explores the target node, the A* algorithm stops searching and begins to backtrack to the center point of the robot's location, displays the planned path on the LCD screen, and leads the tourists to the destination.

2. The navigation path planning method based on artificial intelligence according to claim 1, characterized in that: In step S100, the scanning unit scans the interior environment of the museum. The scanning unit consists of a laser scanner and a camera. The 2D environment construction unit constructs a global 2D environment based on the scanning results of the scanning unit.

3. The navigation path planning method based on artificial intelligence according to claim 1, characterized in that: In step S200, the raster unit rasterizes the 2D composition. The rasterization method is to use Adobe Photoshop to convert the vector graphics and non-pixel images of the 2D composition into pixel images. The embedding unit is used to embed exhibits and locations into the nodes of the raster. The embedding method is to use Adobe Photoshop to create images of exhibits and locations, add them to a new layer, and align them with the nodes.

4. The navigation path planning method based on artificial intelligence according to claim 1, characterized in that: In step S300, the sensor unit senses the time it takes for an object to approach and remain near the robot. When the time exceeds 1.5 seconds, the robot automatically turns its head and body to the side where the object is located, and the communication unit is used to turn on the LCD screen.

5. The navigation path planning method based on artificial intelligence according to claim 1, characterized in that: In step S400, the visualization unit displays the exhibits and locations within the museum to visitors on the screen in the form of a map or a list.

6. The navigation path planning method based on artificial intelligence according to claim 1, characterized in that: In step S500, the selection unit is used for tourists to select the destination they want to reach. The destination includes exhibits and locations within the museum. The sensor detection unit detects the location of the robot. The decision unit plans a path based on the destination selected by the user in the selection unit. The path planning method is as follows: when the tourist selects a specific product, the target point is set at a node next to the exhibit and close to the center of the area; when the tourist selects a location, the target point is set at the node where the entrance of that location is located.

7. The navigation path planning method based on artificial intelligence according to claim 1, characterized in that: In step S600, the A* algorithm unit plans the path according to the evaluation function, selects the fastest path, and ensures that the route is optimal.

8. The navigation path planning method based on artificial intelligence according to claim 1, characterized in that: In step S700, the reporting unit reports and displays the planned path, and the robot immediately leads the tourists to their destination.

Images